Pith. sign in

REVIEW 5 major objections 5 minor 63 references

SynthGuard: Redefining Synthetic Data Generation with a Scalable and Privacy-Preserving Workflow Framework

T0 review · 5 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read SynthGuard claims that synthetic data generation can be run as standardized, auditable workflows entirely within the data owner's domain, producing privacy and utility reports before any data is shared.

desk verdict Useful orchestration work, but the sovereignty guarantee is trust-based, not enforced; the paper overclaims what the framework delivers. read the letter →

arxiv 2507.10489 v1 pith:GROFRCVD submitted 2025-07-14 cs.CR

Eduardo Brito , Mahmoud Shoush , Kristian Tamm , Paula Etti , Liina Kamm This is my paper
classification cs.CR
keywords syntheticdatagenerationprivacy-preservingworkflowssovereigntycomputationalgovernancemeshpipelinespecificationprivacyandutilityevaluationKubeflow
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

SynthGuard is a framework for running synthetic-data generation (SDG) workflows under the data owner's control. The paper argues that by turning an SDG pipeline into a standardized, portable specification that executes inside the owner's own environment, organizations can generate shareable synthetic data without ever transferring raw data. The framework wraps user-chosen generation methods into modular components and adds built-in privacy and utility reports, so the owner can validate outputs before releasing them. Validation against six use cases in two EU projects indicates the design supports on-site, on-premises, and cloud execution, with sublinear scaling up to 100K rows. If the claim holds, regulated sectors can adopt SDG without ceding data sovereignty.

What carries the argument

The central object is the SDG pipeline specification: a declarative, shareable artifact, expressed as an Argo Workflows YAML file, that encodes the directed acyclic graph of modular SDG component tasks, resource allocations, execution settings, and per-stage code. Because the specification is portable and auditable before any sensitive data is loaded, it lets data owners review and approve every step, then execute the identical artifact in their own local, on-prem, or cloud environment. The modular component library and the Kubeflow/Kubernetes orchestration are the machinery that turns this specification into executable, reproducible workflows, while the integrated evaluation components make privacy and utility checks a native part of pipeline execution.

What would settle it

Attach outbound network monitoring to a SynthGuard deployment in its cloud configuration during a 100K-row pipeline run; if any raw data payload leaves the data owner's virtual network, the claim that SDG runs without external data transfer is falsified.

Watch

Extended reading notes

Core claim

The authors propose and implement SynthGuard, an architecture that reorganizes SDG around a declarative pipeline specification rather than around a particular generative model. A pipeline is assembled from modular components (data loading, preprocessing, generation, evaluation) using a Python library and Kubeflow SDK, then exported as an Argo Workflows YAML artifact; the data owner deploys and runs that artifact on a Nix-built, Kubernetes-orchestrated stack inside their own domain. The framework integrates diagnostic, utility, and privacy evaluation (including pMSE, SPECKS, CategoricalCAP, and NewRowSynthesis) so that every run produces reports the owner can review before sharing. The authors state they do not introduce new SDG models; they integrate existing methods such as CTGAN and rule-based generation into governed pipelines. Their validation covers local, on-premises, and cloud deployments, with a scalability benchmark on law-enforcement data showing total runtime growing sublinearly from 1K to 100K rows.

Load-bearing premise

The framework's guarantees rest on the assumption that the requirements drawn from six use cases in two EU projects represent the needs of healthcare, finance, and law enforcement generally.

Editorial extensions

If this is right

  • Data owners can run the whole SDG lifecycle—specification, execution, evaluation, and sharing—without raw data leaving their jurisdiction.
  • Privacy and utility reports are generated automatically per run, giving owners a concrete compliance artifact to review before external sharing.
  • The same pipeline specification is portable across local, on-premises, and cloud environments, supporting multi-cloud or hybrid deployment strategies.
  • SDG methods are interchangeable modules, so switching between CTGAN and rule-based generation does not require rearchitecting the workflow.
  • Standardized pipeline artifacts could become the basis for data-sharing marketplaces where owners retain execution control.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The governance concept could extend beyond SDG to other privacy-enhancing computations, since the pipeline-specification artifact is agnostic to the underlying task.
  • The framework's current single-table scope and the lack of direct privacy/utility evaluation for rule-based synthetic data mean its sovereignty and compliance guarantees are untested for multi-table relational data and for ASD outputs.
  • Regulatory acceptance will determine whether the built-in metrics count as compliance evidence; the paper does not itself argue that any particular regulator accepts these reports.
  • The scalability result is drawn from one law-enforcement dataset; the sublinear trend may not hold for wide tables or data with many categorical features.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

5 major / 5 minor

Summary. The paper presents SynthGuard, a workflow framework for synthetic data generation (SDG) that aims to keep computation and data under the data owner's control. The framework translates user-composed Python pipelines (via the Kubeflow SDK) into portable Argo Workflows specifications, deployed with Nix, Kubernetes (Minikube), and Kubeflow Pipelines. Validation is performed in the context of two EU projects (LAGO and TEADAL), covering six use cases, with a runtime benchmark on three dataset sizes for the law-enforcement use case. The paper claims data sovereignty, compliance verification, scalability, and portability as the framework's principal contributions.

Significance. If the central claims were fully supported, SynthGuard would be a valuable practical contribution: it is an open-source, modular orchestrator that integrates existing SDG and evaluation libraries (e.g., SDV, SynthGauge) into reproducible, auditable pipelines, and it demonstrates a requirements-driven design from real EU projects. The explicit requirement-to-feature mapping (Table 2) and the concrete GitHub artifact are strengths. However, as written, the evidence supports a much weaker claim: the framework can assemble and execute SDG pipelines across three deployment settings, with a preliminary three-point runtime trend. The stronger claims about enforced sovereignty, compliance assurance, and scalability are not established by the current evaluation.

major comments (5)
  1. [§4.2–4.3] The paper's central claim — that SynthGuard enables data owners to 'maintain control over sensitive data and workflows' (Section 1) and 'ensure computational governance' (Abstract) — is not supported by the described implementation. Section 4.3 states that data owners can 'audit and review' the Argo Workflows specification, and Section 4.2 lists 'future support for VM-based TEEs.' No control is described that would prevent a pipeline component from exfiltrating raw data, e.g., through network egress, a shared volume, or log output. Under the current Nix/Kubernetes/Kubeflow stack, the sovereignty guarantee reduces to trusting every component in the specification. Because this guarantee is the paper's claimed novelty, the authors must either implement an enforcement layer (network policies, seccomp/AppArmor, attestation) or explicitly weaken the claim to 'auditable, trust-all-components orchestration.'
  2. [§5 / Table 2 (ALL_R03)] Compliance checking is described as the generation of utility and privacy reports (Table 2, ALL_R03), but Section 5 provides no thresholds, decision rules, or automated gates. For example, the pipeline in Figure 4 produces 'privacy and utility reports' but nothing in the described framework determines whether the synthetic data is actually compliant before release; the abstract's 'ensuring compliance with privacy regulations' is therefore not established. The authors should define measurable acceptance criteria (e.g., pMSE ratio bounds, CAP/TCAP limits) or a documented human-in-the-loop review step with stated responsibilities.
  3. [§3.1 vs §5] The validation is circular in the sense that the requirements in Table 1 were elicited from the LAGO and TEADAL projects (Section 3.1) and the validation in Section 5 consists of showing that SynthGuard satisfies those same requirements (Table 2). This demonstrates internal consistency, not independent confirmation of the framework's broader claims of domain coverage and general compliance support. The 'six distinct use cases' are all drawn from the same two projects, so the claimed representativeness for healthcare, finance, and law enforcement at large is not established. Either add an external use case or reframe the contribution as a requirements-driven design study for these two projects.
  4. [§5 / Fig. 3] The scalability claim rests on Figure 3: three runtime measurements (1K, 10K, and 100K rows) from a single law-enforcement pipeline, with no error bars, no repetitions, and no comparison against another workflow engine or a monolithic baseline. The statement that 'total runtime scaled sublinearly' and that the results 'confirm SynthGuard's scalability' is an overgeneralization from this evidence. Since 'scalable' appears in the title, the authors should either add rigorous benchmarking (multiple pipelines, repeated runs with variance, comparative baselines) or downgrade the claim to a preliminary observation.
  5. [§4.4 / §5] Section 4.4 lists utility and privacy metrics (observed pMSE, standard pMSE, CATCAP, NewRowSynthesis, TCAP, etc.), but Section 5 reports no numerical results for these metrics. Without any reported values, the paper cannot support the claim that SynthGuard 'preserves privacy and utility requirements' (Section 3.2) or that the generated synthetic data is acceptable for sharing. The authors should include representative metric outcomes or clearly state that such evaluations are left to the user.
minor comments (5)
  1. [§1] The text mentions 'the Directive (EU) 2016/860'; the Law Enforcement Directive is (EU) 2016/680, which matches reference [18].
  2. [§5] The phrase 'onlaw enforcementuse case' is missing spaces; it should read 'on the law enforcement use case.'
  3. [§5 / Fig. 3] Figure 3's axes and legend are not described in the text; please clarify whether times are wall-clock or CPU time and note the environment's variance.
  4. [Table 2 vs §5] Table 2 uses 'on-site' in ALL_R01 while Section 5 uses 'on-premises'; please use one term consistently.
  5. [§4.4] The metrics 'standard pMSE' and 'SPECKS' are introduced without definitions; either define them in the text or spell out the acronyms.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity; the framework is validated empirically and its governance claims are architectural, not derived from fitted inputs or self-citation.

full rationale

SynthGuard does not present a mathematical derivation or a fitted parameter renamed as a prediction. Its central contribution is an architectural workflow framework built on standard open-source components (Nix, Kubernetes, Kubeflow, Argo). Section 4.2 explicitly states: 'SynthGuard does not define new SDG models, but rather integrates user-specified methods' — so the SDG results are delegated to external methods, not claimed as outputs of a circular derivation. The runtime measurements in Section 5 (1K/10K/100K rows) are direct empirical benchmarks, not quantities reconstructed from the assumptions. The requirements in Table 1 were elicited from the same LAGO and TEADAL projects used for validation in Table 2, and this self-referential validation is a methodological limitation: it shows the framework satisfies its own specification rather than an independent benchmark. However, that is a weakness in external validity, not circularity under the definitions used here. The requirements are not defined in terms of the validation outcomes, and the framework could fail to meet them (e.g., pipeline execution could time out or fail). No load-bearing assertion relies on a self-citation: references [6], [17], and [31] include overlapping authors but are cited for general background, terminology, or future work, not to force a conclusion. The skeptical concern that sovereignty is not cryptographically enforced (Section 4.2 lists TEEs only as 'future support') is a security limitation, but it does not make the derivation circular. Therefore no circular step is identified.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

SynthGuard rests on domain assumptions about the adequacy of Data Mesh, placement-based privacy, the representativeness of two EU projects, and the validity of standard SDG metrics. It introduces no fitted parameters and no new physical or formal entities.

assumptions (4)
  • domain assumption Data Mesh principles (domain ownership, computational governance, data sovereignty) are appropriate for SDG workflows.
    Section 2 invokes Data Mesh as the architectural foundation without empirical evidence that these principles achieve the claimed governance outcomes in practice.
  • domain assumption Local execution under data owner control guarantees data sovereignty and privacy.
    Section 3.2 states real data stays within the owner's domain, supporting ALL_R01 and ALL_R02, but the paper provides no threat model or data-flow control specification to justify this equivalence.
  • domain assumption The requirements in Table 1, elicited from LAGO and TEADAL, are representative of SDG needs in the claimed sectors.
    Section 3.1 derives requirements from these two projects; the paper assumes this sample covers the solution space without evidence.
  • domain assumption The standard evaluation metrics (pMSE, TCAP, NewRowSynthesis) are valid proxies for privacy and utility.
    Section 4.4 lists metrics from SDV and SynthGauge but does not establish thresholds, calibration, or validation against actual disclosure events.

how reviews work

0 comments
Cite this review

Pith. "Pith review of SynthGuard: Redefining Synthetic Data Generation with a Scalable and Privacy-Preserving Workflow Framework." pith.science (2026). https://pith.science/paper/GROFRCVD

@misc{pith2026250710489,
  author       = {Pith},
  title        = {Pith review of: SynthGuard: Redefining Synthetic Data Generation with a Scalable and Privacy-Preserving Workflow Framework},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/GROFRCVD}},
  note         = {Machine review of arXiv:2507.10489}
}
read the original abstract

The growing reliance on data-driven applications in sectors such as healthcare, finance, and law enforcement underscores the need for secure, privacy-preserving, and scalable mechanisms for data generation and sharing. Synthetic data generation (SDG) has emerged as a promising approach but often relies on centralized or external processing, raising concerns about data sovereignty, domain ownership, and compliance with evolving regulatory standards. To overcome these issues, we introduce SynthGuard, a framework designed to ensure computational governance by enabling data owners to maintain control over SDG workflows. SynthGuard supports modular and privacy-preserving workflows, ensuring secure, auditable, and reproducible execution across diverse environments. In this paper, we demonstrate how SynthGuard addresses the complexities at the intersection of domain-specific needs and scalable SDG by aligning with requirements for data sovereignty and regulatory compliance. Developed iteratively with domain expert input, SynthGuard has been validated through real-world use cases, demonstrating its ability to balance security, privacy, and scalability while ensuring compliance. The evaluation confirms its effectiveness in implementing and executing SDG workflows and integrating privacy and utility assessments across various computational environments.

Figures

Figures reproduced from arXiv: 2507.10489 by the authors.

Figure 1
Figure 1. SDG architectural approach, targeting Data Mesh principles of domain owner￾ship, data sovereignty, and computational governance [PITH_FULL_IMAGE:figures/full_fig_p006_1.png] view at source ↗
Figure 2
Figure 2. SynthGuard framework overview. The pipeline specification is constructed from modular SDG components (left), compiled and orchestrated through Kube￾flow/Kubernetes (centre), and executed under the data owner’s control (right). Outputs include synthetic datasets and their corresponding evaluation reports [PITH_FULL_IMAGE:figures/full_fig_p009_2.png] view at source ↗
Figure 3
Figure 3. Execution time per SDG pipeline component across dataset sizes [PITH_FULL_IMAGE:figures/full_fig_p013_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Top: Example of a law enforcement SDG pipeline visualized via Kubeflow. Bottom: Privacy and utility reports accessible per pipeline component [PITH_FULL_IMAGE:figures/full_fig_p014_4.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

63 extracted references · 61 canonical work pages

  1. [1]

    IEEE Access (2024)

    Alsamhi, S.H., Hawbani, A., Kumar, S., Timilsina, M., Al-Qatf, M., Haque, R., Nashwan, F., Zhao, L., Curry, E.: Empowering Dataspace 4.0: Unveiling Promise of Decentralized Data-Sharing. IEEE Access (2024)

    work page 2024

  2. [2]

    https://argo-workflows.readthedocs.io/en/ (2024), Accessed: 23 September 2024

    Argo Workflows: Overview. https://argo-workflows.readthedocs.io/en/ (2024), Accessed: 23 September 2024

    work page 2024

  3. [3]

    In: CEUR WORKSHOP PROCEEDINGS

    Atzori, M., Ciaramella, A., Diamantini, C., Martino, B., Distefano, S., Facchinetti, T., Montecchiani, F., Nocera, A., Ruffo, G., Trasarti, R., et al.: Dataspaces: Con- cepts, Architectures and Initiatives. In: CEUR WORKSHOP PROCEEDINGS. vol. 3606. CEUR-WS (2024)

    work page 2024

  4. [4]

    Fog Computing: Theory and Practice pp

    Beckman, P., Dongarra, J., Ferrier, N., Fox, G., Moore, T., Reed, D., Beck, M.: Harnessing the computing continuum for programming our world. Fog Computing: Theory and Practice pp. 215–230 (2020)

    work page 2020

  5. [5]

    Building Machine Learn- ing and Deep Learning Models on Google Cloud Platform: A Comprehensive Guide for Beginners pp

    Bisong, E., Bisong, E.: Kubeflow and kubeflow pipelines. Building Machine Learn- ing and Deep Learning Models on Google Cloud Platform: A Comprehensive Guide for Beginners pp. 671–685 (2019)

    work page 2019

  6. [6]

    Proceeding of the 28th International Conference on Enterprise Design, Operations and Computing (2024)

    Brito, E., Castillo, F., Pullonen-Raudvere, P., Werner, S.: TrustOps: Continuously Building Trustworthy Software. Proceeding of the 28th International Conference on Enterprise Design, Operations and Computing (2024)

    work page 2024

  7. [7]

    Data 7(12), 178 (2022) 16 E

    Chandra, G., Siirtola, P., Tamminen, S., Knip, M., Veijola, R., Röning, J.: Impacts of Data Synthesis: A Metric for Quantifiable Data Standards and Performances. Data 7(12), 178 (2022) 16 E. Brito et al

    work page 2022

  8. [8]

    A Survey of Data Synthesis Approaches

    Chang, H., Chen, P., Chou, T., Kao, C., Yu, H., Lin, Y., Chen, Y.: A Survey of Data Synthesis Approaches. CoRRabs/2407.03672 (2024)

    work page Pith review arXiv 2024

Show all 63 references
  1. [9]

    Applied Sciences11(5), 2158 (2021)

    Dankar, F.K., Ibrahim, M.: Fake it till you make it: Guidelines for effective syn- thetic data generation. Applied Sciences11(5), 2158 (2021)

    2021

  2. [10]

    Daskal, J.: Law enforcement access to data across borders: The evolving security and rights issues. J. Nat’l Sec. L. & Pol’y8, 473 (2015)

    2015

  3. [11]

    CoRR abs/2405.20959 (2024)

    Davila, M.F., Groen, S., Panse, F., Wingerath, W.: Navigating Tabular Data Synthesis Research: Understanding User Needs and Tool Capabilities. CoRR abs/2405.20959 (2024)

    2024 arXiv

  4. [12]

    In: Proceedings of the 12th International Conference on Management of Digital EcoSystems

    Dessalk, Y.D., Nikolov, N., Matskin, M., Soylu, A., Roman, D.: Scalable execu- tion of big data workflows using software containers. In: Proceedings of the 12th International Conference on Management of Digital EcoSystems. pp. 76–83 (2020)

    2020

  5. [13]

    Dixon, M.F., Halperin, I., Bilokon, P.: Machine learning in finance, vol. 1170. Springer (2020)

    2020

  6. [14]

    Computers12(10), 198 (2023)

    Donta, P.K., Murturi, I., Casamayor Pujol, V., Sedlak, B., Dustdar, S.: Exploring the potential of distributed computing continuum systems. Computers12(10), 198 (2023)

    2023

  7. [15]

    In: Symposium and Summer School on Service-Oriented Computing

    Eichler, R., Gröger, C., Hoos, E., Schwarz, H., Mitschang, B.: From data asset to data product–the role of the data provider in the enterprise data marketplace. In: Symposium and Summer School on Service-Oriented Computing. pp. 119–138. Springer (2022)

    2022

  8. [16]

    O’Reilly Media (2020)

    El Emam, K., Mosquera, L., Hoptroff, R.: Practical synthetic data generation: balancing privacy and the broad availability of data. O’Reilly Media (2020)

    2020

  9. [17]

    Master’s thesis, Tallinn University of Technology, Tallinn, Estonia (2024)

    Etti, P.: Exploring the use of synthetic data in the public sector: a framework and case study based on the example of the Estonian Police and Border Guard Board. Master’s thesis, Tallinn University of Technology, Tallinn, Estonia (2024)

    2024

  10. [18]

    Official Journal L 11959, 89–131 (2016)

    European Commission: Directive (EU) 2016/680 of the European Parliament and of the Council of 27 April 2016. Official Journal L 11959, 89–131 (2016)

    2016

  11. [19]

    European Commission: Art. 22 GDPR. Automated individual decision-making, including profiling. Intersoft Consulting, https://gdpr-info. eu/art-22-gdpr (2020)

    2020

  12. [20]

    European Commission: Data Governance Act (2022), https:// digital-strategy.ec.europa.eu/en/policies/data-governance-act

    2022

  13. [21]

    eu/en/policies/data-act

    European Commission: Data Act (2023),https://digital-strategy.ec.europa. eu/en/policies/data-act

    2023

  14. [22]

    Sin- gle Resolution Board v European Data Protection Supervisor

    General Court (Eighth Chamber, Extended Composition): Judgment of the Gen- eral Court (Eighth Chamber, Extended Composition) of 26 April 2023. Sin- gle Resolution Board v European Data Protection Supervisor. (4 2023),https: //eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX%3...

    2023

  15. [23]

    Internet Things19, 100514 (2022)

    Gill, S.S., Xu, M., Ottaviani, C., Patros, P., Bahsoon, R., Shaghaghi, A., Golec, M., Stankovski, V., Wu, H., Abraham, A., Singh, M., Mehta, H., Ghosh, S.K., Baker, T., Parlikad, A.K., Lutfiyya, H., Kanhere, S.S., Sakellariou, R., Dustdar, S., Rana, O.F., Brandic, I., Uhlig, S...

    2022

  16. [24]

    ACM Computing Surveys 57(1), 1–36 (2024)

    Goedegebuure, A., Kumara, I., Driessen, S., Van Den Heuvel, W.J., Monsieur, G., Tamburri, D.A., Nucci, D.D.: Data mesh: a systematic gray literature review. ACM Computing Surveys 57(1), 1–36 (2024)

    2024

  17. [25]

    Deploying AI in the Enterprise: IT Approaches for Design, DevOps, Governance, Change Management, Blockchain, and Quantum Computing pp

    Hechler, E., Oberhofer, M., Schaeck, T., Hechler, E., Oberhofer, M., Schaeck, T.: The operationalization of AI. Deploying AI in the Enterprise: IT Approaches for Design, DevOps, Governance, Change Management, Blockchain, and Quantum Computing pp. 115–140 (2020) SynthGuard 17

    2020

  18. [26]

    In: ARES

    Hittmeir, M., Ekelhart, A., Mayer, R.: On the Utility of Synthetic Data: An Em- pirical Evaluation on Machine Learning Tasks. In: ARES. pp. 29:1–29:6. ACM (2019)

    2019

  19. [27]

    CoRR abs/1710.08874 (2017)

    Howe, B., Stoyanovich, J., Ping, H., Herman, B., Gee, M.: Synthetic Data for Social Good. CoRR abs/1710.08874 (2017)

    2017 arXiv

  20. [28]

    Hu, Y., Wu, F., Li, Q., Long, Y., Garrido, G.M., Ge, C., Ding, B., Forsyth, D.A., Li, B., Song, D.: SoK: Privacy-Preserving Data Synthesis. In: SP. pp. 4696–4713. IEEE (2024)

    2024

  21. [29]

    Big Data & Society8(1), 2053951720982012 (2021)

    Hummel, P., Braun, M., Tretter, M., Dabrock, P.: Data sovereignty: A review. Big Data & Society8(1), 2053951720982012 (2021)

    2021

  22. [30]

    Science 349(6245), 255–260 (2015)

    Jordan, M.I., Mitchell, T.M.: Machine learning: Trends, perspectives, and prospects. Science 349(6245), 255–260 (2015)

    2015

  23. [31]

    In: Privacy and Identity Management

    Kamm, L., Bogdanov, D., Brito, E., Ostrak, A.: Blueprints for Deploying Privacy Enhancing Technologies in E-Government. In: Privacy and Identity Management. IFIP Advances in Information and Communication Technology, vol. 695, pp. 3–19. Springer (2023)

    2023

  24. [32]

    Kononenko, I.: Machine learning for medical diagnosis: history, state of the art and perspective. Artif. Intell. Medicine23(1), 89–109 (2001)

    2001

  25. [33]

    https://www.kubeflow.org/docs/components/ pipelines/ (2024), Accessed: 23 September 2024

    Kubeflow: Pipelines Overview. https://www.kubeflow.org/docs/components/ pipelines/ (2024), Accessed: 23 September 2024

    2024

  26. [34]

    https://kubernetes.io/ (2024), Accessed: 15 October 2024

    Kubernetes: Docs. https://kubernetes.io/ (2024), Accessed: 15 October 2024

    2024

  27. [35]

    In: Privacy and Identity Management

    Kwatra, S., Torra, V.: Empirical Evaluation of Synthetic Data Created by Genera- tive Models via Attribute Inference Attack. In: Privacy and Identity Management. IFIP Advances in Information and Communication Technology, vol. 695, pp. 282–

  28. [36]

    Lenatti,M.,Paglialonga,A.,Orani,V.,Ferretti,M.,Mongelli,M.:Characterization of Synthetic Health Data Using Rule-Based Artificial Intelligence Models. IEEE J. Biomed. Health Informatics27(8), 3760–3769 (2023)

    2023

  29. [37]

    Procedia Computer Science196, 263–271 (2022)

    Machado, I.A., Costa, C., Santos, M.Y.: Data mesh: concepts and principles of a paradigm shift in data architectures. Procedia Computer Science196, 263–271 (2022)

    2022

  30. [38]

    Majeed, A.: Attribute-Centric and Synthetic Data Based Privacy Preserving Meth- ods: A Systematic Review. J. Cybersecur. Priv.3(3), 638–661 (2023)

    2023

  31. [39]

    In: Proceedings of the 2024 ACM/IEEE 44th Interna- tional Conference on Software Engineering: New Ideas and Emerging Results

    Malka, J., Zacchiroli, S., Zimmermann, T.: Reproducibility of Build Environments through Space and Time. In: Proceedings of the 2024 ACM/IEEE 44th Interna- tional Conference on Software Engineering: New Ideas and Emerging Results. pp. 97–101 (2024)

    2024

  32. [40]

    https://minikube.sigs.k8s.io/docs/ (2024), Accessed: 12 November 2024

    Minikube: Overview. https://minikube.sigs.k8s.io/docs/ (2024), Accessed: 12 November 2024

    2024

  33. [41]

    In: 2019 Amity international conference on artificial intelligence (AICAI)

    Nagpal, A., Gabrani, G.: Python for data analytics, scientific and technical appli- cations. In: 2019 Amity international conference on artificial intelligence (AICAI). pp. 140–145. IEEE (2019)

    2019

  34. [42]

    Nikolenko, S.I.: Synthetic data for deep learning, vol. 174. Springer (2021)

    2021

  35. [43]

    Nix: Declarative builds and deployments.https://nixos.org/ (2024), Accessed: 5 September 2024

    2024

  36. [44]

    IEEE Access9, 68028–68043 (2021)

    Oleghe, O.: Container placement and migration in edge computing: Concept and scheduling models. IEEE Access9, 68028–68043 (2021)

    2021

  37. [45]

    IEEE Access12, 88048–88074 (2024) 18 E

    Osorio-Marulanda, P.A., Epelde, G., Hernandez, M., Isasa, I., Reyes, N.M., Iraola, A.B.: Privacy Mechanisms and Evaluation Metrics for Synthetic Data Generation: A Systematic Review. IEEE Access12, 88048–88074 (2024) 18 E. Brito et al

    2024

  38. [46]

    Park, N., Mohammadi, M., Gorde, K., Jajodia, S., Park, H., Kim, Y.: Data Syn- thesis based on Generative Adversarial Networks. Proc. VLDB Endow.11(10), 1071–1083 (2018)

    2018

  39. [47]

    In: DSAA

    Patki, N., Wedge, R., Veeramachaneni, K.: The Synthetic Data Vault. In: DSAA. pp. 399–410. IEEE (2016)

    2016

  40. [48]

    Computational and structural biotechnology journal (2024)

    Pezoulas, V.C., Zaridis, D.I., Mylona, E., Androutsos, C., Apostolidis, K., Tachos, N.S., Fotiadis, D.I.: Synthetic data generation methods in healthcare: A review on open-source tools and methods. Computational and structural biotechnology journal (2024)

    2024

  41. [49]

    In: SSDBM

    Ping, H., Stoyanovich, J., Howe, B.: DataSynthesizer: Privacy-Preserving Synthetic Datasets. In: SSDBM. pp. 42:1–42:5. ACM (2017)

    2017

  42. [50]

    In: International Conference on Enterprise Design, Operations, and Computing

    Rambert, P., Rychkova, I.: Implications of Trust in Cyber-Physical Systems De- sign: The ASSA Case Study. In: International Conference on Enterprise Design, Operations, and Computing. pp. 201–218. Springer (2024)

    2024

  43. [51]

    Knowledge-Based Systems194, 105596 (2020)

    Sengupta, S., Basak, S., Saikia, P., Paul, S., Tsalavoutis, V., Atiah, F., Ravi, V., Peters, A.: A review of deep learning with special emphasis on architectures, ap- plications and recent trends. Knowledge-Based Systems194, 105596 (2020)

    2020

  44. [52]

    Fron- tiers Artif

    Shi, J., Wang, D., Tesei, G., Norgeot, B.: Generating high-fidelity privacy-conscious synthetic patient data for causal effect estimation with multiple treatments. Fron- tiers Artif. Intell.5 (2022)

    2022

  45. [53]

    IEEE Transactions on Software Engineering48(9), 3361–3377 (2021)

    Shi, Y., Wen, M., Cogo, F.R., Chen, B., Jiang, Z.M.: An experience report on producing verifiable builds for large-scale commercial systems. IEEE Transactions on Software Engineering48(9), 3361–3377 (2021)

    2021

  46. [54]

    BMC medical research methodology19, 1–18 (2019)

    Sidey-Gibbons, J.A., Sidey-Gibbons, C.J.: Machine learning in medicine: a practi- cal introduction. BMC medical research methodology19, 1–18 (2019)

    2019

  47. [55]

    Journal of Artificial Intelligence Research and Applications1(2), 292–332 (2021)

    Singh, J.: The Rise of Synthetic Data: Enhancing AI and Machine Learning Model Training to Address Data Scarcity and Mitigate Privacy Risks. Journal of Artificial Intelligence Research and Applications1(2), 292–332 (2021)

    2021

  48. [56]

    Soltana, G., Sabetzadeh, M., Briand, L.C.: Synthetic data generation for statistical testing. In: ASE. pp. 872–882. IEEE Computer Society (2017)

    2017

  49. [57]

    In: ICLR

    Stoian, M.C., Dyrmishi, S., Cordy, M., Lukasiewicz, T., Giunchiglia, E.: How Re- alistic Is Your Synthetic Data? Constraining Deep Generative Models for Tabular Data. In: ICLR. OpenReview.net (2024)

    2024

  50. [58]

    Journal of Artificial Intelligence Research and Applications1(2), 58–88 (2021)

    Tatineni, S., Boppana, V.R.: AI-Powered DevOps and MLOps Frameworks: En- hancingCollaboration,Automation,andScalabilityinMachineLearningPipelines. Journal of Artificial Intelligence Research and Applications1(2), 58–88 (2021)

    2021

  51. [59]

    In: Privacy in Statistical Databases: UNESCO Chair in Data Privacy, International Conference, PSD 2018, Valencia, Spain, September 26–28, 2018, Proceedings

    Taub, J., Elliot, M., Pampaka, M., Smith, D.: Differential correct attribution prob- ability for synthetic data: an exploration. In: Privacy in Statistical Databases: UNESCO Chair in Data Privacy, International Conference, PSD 2018, Valencia, Spain, September 26–28, 2018, Proc...

    2018

  52. [60]

    Data9(1), 14 (2024)

    Tolas, R., Portase, R., Potolea, R.: GeMSyD: Generic Framework for Synthetic Data Generation. Data9(1), 14 (2024)

    2024

  53. [61]

    Available at SSRN 3762686 (2020)

    Wagner, P.: Privacy Enhancing Technologies and Synthetic Data. Available at SSRN 3762686 (2020)

    2020

  54. [62]

    In:2020Internationalconferenceonartificialintelligenceandcomputerengineering (ICAICE)

    Zhou, Y., Yu, Y., Ding, B.: Towards MLops: A case study of ML pipeline platform. In:2020Internationalconferenceonartificialintelligenceandcomputerengineering (ICAICE). pp. 494–500. IEEE (2020)

    2020

  55. [63]

    Zielke, T.: Is artificial intelligence ready for standardization? In: Systems, Software andServicesProcessImprovement:27thEuropeanConference,EuroSPI2020,Düs- seldorf, Germany, September 9–11, 2020, Proceedings 27. pp. 259–274. Springer (2020)

    2020

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.